OpenAI has introduced product feed ads to its Ads Manager beta. Retail advertisers can now upload their full product catalogues and generate ads directly from feed data, matching individual items to users mid-conversation, without building campaigns product by product.
According to OpenAI, feed-based ad formats have been among the strongest performers in its Ads Beta to date. The move extends its existing advertising infrastructure toward a model that Google and Meta have used to scale retail advertising for years: dynamic product ads that match inventory to intent automatically.
The difference is where those ads are served. ChatGPT conversations, particularly purchase-focused ones, represent a different kind of intent signal than a Google search query or a Meta scroll. The user isn’t browsing. They’re asking. The query's specificity is higher, and the context around it is richer. Whether that translates to commercial performance at scale is the question the beta will answer.
What’s Actually Launched
What we know on June 11, 2026
- Format: Product feed ads, catalogue uploaded, ads generated dynamically per item
- Platform: ChatGPT, via OpenAI Ads Manager beta
- Access: Beta — retail advertisers can review feed requirements and begin campaigns from uploaded catalogues
- Targeting mechanism: Matches products to purchase-intent conversations — specifics of the matching model not yet published
- First reported: Menachem Ani, Founder of JXT Group, via X — shared the OpenAI beta email
The mechanics mirror what Google Shopping and Meta’s Dynamic Product Ads do: advertisers provide a structured feed, the platform handles ad generation and matching. The operational lift for a retailer already running feed-based campaigns on other channels is relatively low. The catalogue structure required is familiar.
Why This Matters for eCommerce
ChatGPT’s user base is now well into the hundreds of millions of active users.
A meaningful share of those conversations are product research, comparison, and purchase queries. Until now, OpenAI’s advertising offering was limited enough that most retail advertisers were watching rather than testing. Product feed support changes that calculus.
The intent quality argument for ChatGPT ads is genuine. When a user asks “what’s the best running shoe for wide feet under £100,” the query contains intent signals, constraint signals, and context that a standard search query rarely captures in one place. Matching a product feed to that conversation, if the matching model is well-built, could produce more relevant ad placements than the same query expressed as a keyword search.
That’s the upside case. The unknowns are significant.
Feed-based ads are only as good as the feed. Before evaluating ChatGPT as a channel, the same question applies here as everywhere else: are your product titles, attributes, pricing, and inventory accurate? A dynamic ad system amplifies feed quality in both directions.
How This Sits Against Google and Meta
Google Shopping / PMax
Feed-based. Mature auction. Deep performance data. Intent from search query. Scale and infrastructure that ChatGPT doesn't have yet. The benchmark everything else gets measured against.
Meta Dynamic Product Ads
Feed-based. Intent inferred from behaviour rather than expressed in a query. Retargeting strength. Reach across Facebook and Instagram. Discovery-led rather than demand-capture.
ChatGPT Product Feed Ads
Feed-based. Intent expressed in natural language — richer context than a keyword, different from a search query. Conversational format. Beta stage. Performance data not yet publicly available.
The Honest Question
Does conversational intent convert differently from search intent? And does OpenAI's matching model have the commercial signal depth that Google and Meta have built over fifteen years? We don't know yet.
What’s Missing From the Announcement
OpenAI hasn’t published the technical specifications of how product matching works inside ChatGPT conversations. The auction mechanics, bid structure, reporting granularity, and attribution model are all beta-stage and not yet public. For brands evaluating whether to join the beta, those details matter, particularly attribution, where ChatGPT’s position in a purchase journey relative to Google and Meta needs to be understood before spend is allocated.
The “strongest-performing format in beta” claim from OpenAI is directionally interesting and worth treating with the standard caution applied to any platform’s self-reported performance data from early testing. Sample sizes are small, the advertiser mix is curated, and the novelty effect on engagement hasn't been controlled for.
What To Do With This Now
For most e-commerce brands, this is a watch-and-test moment rather than a pivot. The infrastructure question is low friction: if you’re already running Google Shopping or Meta DPA, you have the feed structure ChatGPT requires. The incremental cost of testing is manageable.
The more important question is measurement. Before adding ChatGPT as a channel, the attribution framework needs to be in place. How will you understand what ChatGPT ads are contributing versus Google and Meta and whether the customers they bring in are incremental or duplicating demand you were already capturing elsewhere?
- Audit your feed quality first: Title structure, attribute completeness, pricing accuracy. Dynamic ads surface feed problems at scale.
- Define your incrementality test before you start: Without a measurement framework, you're spending blind.
- Don’t reallocate Google or Meta budget to test it: Run ChatGPT as an additive channel with a defined test budget and evaluation period.
- Watch the beta closely: Feed requirements, bid structure, and reporting will develop quickly. What’s true in June may look different by Q3.
360 OM View
This is a significant structural move from OpenAI. Product feed ads are the commercial infrastructure that turns a conversational platform into a retail advertising channel. The intent quality case is real. The performance data isn't yet. Test it carefully, measure it properly, and don't let the announcement cycle push you into budget decisions before the beta has produced enough signal to justify them.
Need a fresh perspective? Let’s talk.
At 360 OM, we specialise in helping businesses take their marketing efforts to the next level. Our team stays on top of industry trends, uses data-informed decisions to maximise your ROI, and provides full transparency through comprehensive reports.
.png)
